Style Transfer Generator for Dataset Testing Classification
Open Access
- 1 April 2022
- journal article
- Published by Politeknik Ganesha in sinkron
- Vol. 7 (2), 448-454
- https://doi.org/10.33395/sinkron.v7i2.11375
Abstract
The development of the Generative Adversarial Network is currently very fast. First introduced by Ian Goodfellow in 2014, its development has accelerated since 2018. Currently, the need for datasets is sometimes still lacking, while public datasets are sometimes still lacking in number. This study tries to add an image dataset for supervised learning purposes. However, the dataset that will be studied is a unique dataset, not a dataset from the camera. But the image dataset by doing the augmented process by generating from the existing image. By adding a few changes to the augmentation process. So that the image datasets become diverse, not only datasets from camera photos but datasets that are carried out with an augmented process. Camera photos added with painting images will become still images with a newer style. There are many studies on Style transfer to produce images in drawing art, but it is possible to generate images for the needs of image datasets. The resulting force transfer image data set was used as the test data set for the Convolutional Neural Network classification. Classification can also be used to detect specific objects or images. The image dataset resulting from the style transfer is used for the classification of goods transporting vehicles or trucks. Detection trucks are very useful in the transportation system, where currently many trucks are modified to avoid road feesKeywords
This publication has 14 references indexed in Scilit:
- Mutual stain conversion between Giemsa and Papanicolaou in cytological images using cycle generative adversarial networkHeliyon, 2021
- An intelligent electromagnetic environment reconstruction method based on super-resolution generative adversarial networkPhysical Communication, 2020
- Coefficients of non-negative quasi-Cartan matrices, their symmetrizers and Gram matricesDiscrete Applied Mathematics, 2020
- Optimising realism of synthetic images using cycle generative adversarial networks for improved part segmentationComputers and Electronics in Agriculture, 2020
- Kiwifruit detection in field images using Faster R-CNN with VGG16IFAC-PapersOnLine, 2019
- Arbitrary Style Transfer in Real-Time with Adaptive Instance NormalizationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2017
- Improved Texture Networks: Maximizing Quality and Diversity in Feed-Forward Stylization and Texture SynthesisPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2017
- Controlling the least eigenvalue of a random Gram matrixLinear Algebra and its Applications, 2016
- A Neural Algorithm of Artistic StyleJournal of Vision, 2016
- Gram matrix of Bernstein basis: Properties and applicationsJournal of Computational and Applied Mathematics, 2015